Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Suspicious data pattern in recent Venezuelan election

statmodeling.stat.columbia.edu

Post reply on HN
Print view View original

371–380 of 536 posts

Re: Suspicious data pattern in recent Venezuelan election

#371

Post by foobarqux »

lolinder wrote: post #365
foobarqux wrote: post #354

Earlier quoted context omitted.

Because maybe they didn't have direct access to the original numbers? You can't understand that some consumer of the data might not get the original CSV only some summary?

A summary that includes the exact number of votes cast and the percent for each candidate but not the breakdown by candidate? Why would someone go out of their way to construct a CSV that has the tally by candidate removed* but still has the total vote count? What would that CSV even look like? * Yes, the tally would have to be removed, because presumably there's a spreadsheet somewhere that was used to generate perc…

Yes, the percentages are the most important number, the number everyone is interested in. The next most important number is the voter turnout. You can verify this by looking at the newspaper headlines of any election. Again unless you are an electioneer no one cares about the raw numbers so it would not be surprising that only the percentages and total are communicated to the public relations department.

I don't understand your comments about the CSV: I'm saying that the raw CSV is not being distributed, only the summary statistics.

Re: Suspicious data pattern in recent Venezuelan election

#372

Post by 9cb14c1ec0 »

I find it quite amusing to watch people declare the Venezuela election stolen by the very same metrics they loudly proclaimed meant absolutely nothing in the 2020 US elections. At this point, it's not a hard bet to think I'll have more mirth coming my way in November.

Re: Suspicious data pattern in recent Venezuelan election

#373

Post by culi »

olalonde wrote: post #306
culi wrote: post #298

Earlier quoted context omitted.

Depends which polls you're looking at. So far, no non-US-linked pollsters have shown Maduro behind (that is, behind the US backed opposition candidate). If you look at something like Hinterlaces you get a very different picture

According to this Venezuelan news article[0], several did: > Well-known pollsters in the Venezuelan political sphere, such as Datanálisis, Datincorp, Delphos, and Consultores 21, along with the emerging Poder y Estrategia, indicated that Urrutia had more than 50% of voting intentions. [0] https://www.lapatilla.com/2024/07/15/la-guerra-de-encuestas-...

Thanks for the article. I had to translate as my Spanish isn't very good. These seem to be the main points

> Surveys of career-marking firms in the Venezuelan political environment, such as Datanalisis, Datincorp, Delphos and Consultores 21, as well as the incipient Power and Strategy, by political scientist Ricardo Ríos, claim that the postulate of anti-chavismo accumulates more than 50 percent of the intention to vote.

> However, others, such as Hinterlaces and some practically unknown in the Venezuelan or recent data market, including Polymarket, BMI Orientation and DataViva, conclude that Maduro leads his surveys with between 54 percent and 70 percent of the preference. Another, ECSC, talks about a technical tie tipped slightly towards opposition.

> The C-INFORMA Information Coalition, made up of media teams such as Media-analysis, Cocuyo Effect, Fake News Hunters and Probox, concluded this month that 6 out of 14 firms evaluated, recently created and dubious credibility, have published 37 public opinion studies - used in a strategy - to manipulate the country's.

The main point of the article seems to be that there's a lot of misleading polling in general. Which is a fair point, but does kinda leave us outsiders in the dark

Re: Suspicious data pattern in recent Venezuelan election

#374

Post by glenstein »

ARandomerDude wrote: post #311
berdario wrote: post #215

Earlier quoted context omitted.

I was curious to check the National Electoral Council website, to get their latest tally straight from the source... http://www.cne.gov.ve/ And the website is currently not loading. They might simply be struggling to keep up with traffic, but being under some kind of "cyber attack" is also plausible

This would also be fairly trivial to fake if you wanted to appear to be under attack.

Yeah but I think this is the point where you have to take the pro-evidence side. Speculation with increasingly complex assumptions have to overcome a default assumption in favor of simplicity, and it requires evidence to overcome that default assumption.

In this case I would say the simpler assumptions are just that they did steal the election and that the website is down due to global traffic.

Re: Suspicious data pattern in recent Venezuelan election

#375

Post by lolinder »

foobarqux wrote: post #371
lolinder wrote: post #365

Earlier quoted context omitted.

A summary that includes the exact number of votes cast and the percent for each candidate but not the breakdown by candidate? Why would someone go out of their way to construct a CSV that has the tally by candidate removed* but still has the total vote count? What would that CSV even look like? * Yes, the tally would have to be removed, because presumably there's a spreadsheet somewhere that was used to generate perc…

Yes, the percentages are the most important number, the number everyone is interested in. The next most important number is the voter turnout. You can verify this by looking at the newspaper headlines of any election. Again unless you are an electioneer no one cares about the raw numbers so it would not be surprising that only the percentages and total are communicated to the public relations department. I don't unde…

> Yes, the percentages are the most important number, the number everyone is interested in.

Then why did the hypothetical sub-sub-librarian who put together the final spreadsheet feel the need to go back and repopulate those numbers? Clearly they thought people would want to see them, right?

> The next most important number is the voter turnout. You can verify this by looking at the newspaper headlines of any election.

So, in this hypothetical, when the tallies per candidate are expressed as percentages it's because percentages are the natural way to think about these things, but when voter turnout is expressed in raw numbers that's because raw numbers are the natural way to think about voter turnout?

Voter turnout is the only number in the set that I could possibly see making sense to express only as a percentage!

Re: Suspicious data pattern in recent Venezuelan election

#377

Post by culi »

berdario wrote: post #333
culi wrote: post #287

Earlier quoted context omitted.

Exactly my point, thanks! This form of analysis has been used to verify the validity of past election results. All were within 2 points of forecasts based on this data. We still don't have enough results to do this type of analysis yet, but we surely will eventually. The group that did the analysis you're linking is AltaVista. They are linked with the opposition, but their same analysis validated past results. Their…

https://resultadosconvzla.com/ Claims Actas digitalizadas: 24.576 (81,85%) > Their current analysis obviously doesn't but they also admit that their sample is biased towards anti-Maduro centers. If that's true, 81% of the total would already be quite representative (and less subject to be biased by anti-Maduro constituencies being overrrepresented)

These are forecasts based on pictures of ballot tallies. And this website is made by the opposition. I'm not saying they're wrong, I just think it's silly to take their word for it instead of just waiting until we have more data and these results have been looked over by independent parties

Re: Suspicious data pattern in recent Venezuelan election

#378

Post by nequo »

culi wrote: post #254
kergonath wrote: post #136

Earlier quoted context omitted.

> Gee I wonder why he might be paranoid about foreign operatives trying to violently oust him Because he’s an unpopular dictator who knows fairly well for having seen it in the recent past that you can go very quickly from getting 99% of the vote in an election to dead in a ditch? In his position you’re either paranoid or dead. This does not make him a better person.

You can have that personal belief, but independent analyses of past election results have upheld their validity. If he's a dictator, he's one that's earned the popular vote of Venezuela multiple times. Perhaps this time it's different

> but independent analyses of past election results have upheld their validity

What do you consider to be independent? I’m seeing a lot of the opposite here for 2018 for example:

https://en.wikipedia.org/wiki/2018_Venezuelan_presidential_e...

Re: Suspicious data pattern in recent Venezuelan election

#379

Post by glenstein »

gmaster1440 wrote: post #69
dinobones wrote: post #22

Here's this re-explained with a simpler example. Imagine you have 1,000 votes. You want to show that your political party got 60% of the vote, so, you claim: My party: 600 votes Opposition: 300 votes Other: 100 votes Presto, we got a good breakdown. The people will buy it.... It makes sense that 600 is exactly 60% of 1,000, because this was an artificial example. But in the real world, we don't get 1,000 votes. We ge…

If I'm in charge of forging a presidential election, how difficult is it for me to use realistic, "ugly" numbers to sell it more effectively?

Actually think this is a fascinating question, although perhaps for different reasons within whatever reasons might have led you to ask it.

I think you raise a legitimate point that it's not that hard to create ugly numbers.

I also think that authoritarian social dynamics come to these questions with a kind of brutal simplicity, lack of intellectual curiosity or creativity, and a lot of the traits that would entail a value for democracy are mutually exclusive with the brute simplicity of authoritarian mindset. And so the story they choose to tell of how they won is going to have similar hallmarks of brute simplicity and absence of nuance.

Re: Suspicious data pattern in recent Venezuelan election

#380

Post by selimthegrim »

immibis wrote: post #345
stouset wrote: post #241

Earlier quoted context omitted.

Unless done carefully this will almost certainly fail Benford’s Law. Manipulating statistics is harder than you think.

The Biden election in 2020 also failed Benford's law - unless you're suggesting that one was fake, it seems that failing Benford's law is okay.

There was a good paper in an American Statistical Association journal about this.

https://chance.amstat.org/2022/04/benfords-law-votes/

https://www.tandfonline.com/doi/abs/10.1080/09332480.2022.20...

Post reply on HN